Overview

The AI Duplicate Detection & Contact Merging feature helps you identify potential duplicate contacts and merge them into a single record.

Using intelligent AI duplicate detection, the system reviews contact data and suggests possible matches so you can decide what should be merged and what should remain separate.

Availability of this feature depends on your plan. Contact your account manager to learn more.
 

Where to Find It

From the Contacts module, go to Data Hygiene and click Review Duplicates.

Note: Data Hygiene was previously called Data Update Verification.

Review Duplicates

The Review Duplicates page displays potential duplicate contacts for review.

If there are no duplicates to review yet, the page shows an empty state. The system continues checking for duplicates in the background.

Choose Contacts to Merge

Click Review to open a potential duplicate group. Then select the contacts that represent the same person and should be merged. If the records should not be combined, click Keep Separate.

If you click Keep Separate, that merge suggestion is permanently hidden. The system will not suggest merging those specific contacts again, but they can still be merged manually later.

Choose What to Keep

After selecting the contacts to merge, the system highlights the fields with different values.

Review each field and choose which information to keep. Your selections determine the final merged record.

Automatic Scanning

Glue Up Copilot AI automatically scans for potential duplicates every two weeks.

You can hover over the AI indicator in the Review Duplicates tab to see this information.

Notifications

When Glue Up Copilot AI detects potential duplicates, it sends a notification with suggested merges. The notification displays the number of identified duplicates.

Admins are notified both:

• In-app via notifications

• By email

How Duplicate Detection Works

The system detects overlapping contacts by analyzing:

• Names
• Email addresses
• Company details

It uses typo detection, nickname recognition, and weighted similarity scoring to calculate a Match Score. This score helps you decide whether records should be merged or kept separate.